The Last Week to Change Your Ad Account Before Peak

Black Friday falls on 27 November 2026, and the last comfortable week to make structural changes to your ad account is the week commencing 12 October. After that, every change you make is competing with the learning it resets, in the weeks when learning is most expensive.

This is the least glamorous deadline of the quarter and the one that costs the most to miss. This post covers what counts as a structural change, why the timing is fixed rather than a preference, and what to do if you are already past it.

Why does mid October matter more than mid November?

Because campaigns need roughly two weeks to exit the learning phase, and peak trading starts six weeks before Black Friday rather than on it. Counting back from 27 November, six weeks out is the week commencing 12 October. A campaign launched that week has time to learn through late October, stabilise, and be a known quantity by the time budgets step up.

A campaign launched in mid November is still learning during Cyber 5. It will spend your most expensive impressions gathering the data that a mid October launch already has, and it will do it while the auction is at its peak. Across the Webtopia portfolio in Q4 2025, cost per acquisition ran $25.23 in mid October and $50.16 by mid December. Learning in October costs roughly half what learning in December costs.

What counts as a structural change?

Anything that alters how the campaign is set up rather than what it is spending. Targeting changes, campaign or ad set restructures, new conversion events, optimisation goal changes, attribution setting changes, and new campaign launches all reset learning even though historical data is retained.

Things that do not reset learning include budget adjustments within sensible increments, adding new creative into an existing ad set, and switching offers on product ads through catalogue overlays. That last one is worth knowing, because it means you can change your promotional message through the season without paying the learning cost, which is exactly what the phase to phase pivots of peak trading require.

What should you change now, before the window closes?

Five things, and targeting guardrails are the one most often missed.

Open your demographic guardrails. Purchase intent at Black Friday strengthens with age, and tightly age capped strategies exclude some of the audiences most likely to buy. Open targeting from early October and keep it open to mid January, then reapply your normal rules afterwards. Set up your catalogue overlays now so promotional messaging can change later without a structural edit. Launch any new campaign you intend to run at peak, so it learns on cheap impressions. Confirm your conversion event and optimisation goal are the ones you want to trade against for the whole quarter. And verify tracking end to end, including server side events, because fixing a tracking problem in November means fixing it during the learning phase you just reset.

What if you have already missed the window?

Change less, and change earlier in the day rather than later in the month. The principle is that each structural change buys a fixed cost in learning, so the question becomes which changes are worth paying for.

A tracking fix is almost always worth it, because broken measurement compounds across the whole quarter. A tempting new campaign structure usually is not, because you are trading a known quantity for an unknown one at the worst possible moment. If a change is genuinely necessary in November, make it at the start of a week rather than mid flight, and accept that the campaign will be unreliable for the following fortnight.

Should you launch new campaigns cold during Q4?

No. Launching cold in Q4 means introducing an unproven campaign into the most competitive auction of the year, and it is the single most avoidable mistake we see at this point in the calendar.

The alternative is the testing funnel. Launch a wide set of creative into a dedicated testing campaign, promote the emerging winners into your business as usual campaigns, switch them off in testing, and let the next crop surface. The structure stays stable while the creative inside it changes, which gives you novelty without resetting anything.

How does this connect to budget?

Directly, because a campaign that is still learning cannot absorb a budget increase cleanly. Scaling is capped at roughly 30% per seven days before performance degrades, and that cap assumes the campaign is out of learning to begin with.

The practical sequence is stable structure first, then budget. Get the account into the shape you want it in October, let it learn, then use November to scale what is already optimised. We cover the scaling maths and how far back you need to start in Q4 budget planning and pacing.

What comes after this deadline

This deadline is the first of seven phases in the Peak Trading Calendar, which runs from six weeks before Black Friday through January and sets out the media, creative, offer, website and measurement decision for each phase.

The calendar sits in the Q4 Profit Playbook, alongside the five operating rules that frame it, the demand timing data for four categories, and the readiness scorecard that finds the gaps before the season does. It is free and built for founder led DTC brands at $5M to $30M. Most teams keep the calendar chapter open from October onward.

If you want the account itself pressure tested before the window shuts, Webtopia does that as an ecommerce paid media agency working with founder led brands, or you can book a Q4 growth review.

Frequently asked questions

When is Black Friday 2026?

Friday 27 November 2026, with Cyber Monday on 30 November. Six weeks before Black Friday is the week commencing 12 October, which is the practical start of peak trading.

How long does the Meta learning phase take?

Roughly two weeks for most campaigns. That is why a campaign intended to carry peak budget should be launched and learning by mid October rather than in November.

Does changing the budget reset learning?

Sensible budget adjustments do not, but large sudden increases damage performance regardless. Scaling by no more than around 30% per seven days is the working rule.

Can I change my offer during peak without resetting learning?

Yes, using catalogue overlays on product ads. Setting those up in October is what makes the phase to phase offer pivots possible later without a structural edit.

Should I tighten targeting for Black Friday?

The opposite. Purchase intent at Black Friday strengthens with age, so open demographic guardrails from early October through to mid January and reapply your usual rules afterwards.